From hard refusals to safe-completions: toward output-centric safety training
核心要点速览
Discover how OpenAI's new safe-completions approach in GPT-5 improves both safety and helpfulness in AI responses—moving beyond hard refusals to nuanced, output-centric safety training for handling dual-use prompts.
查看完整报道与分享
完整图文、背景资料与讨论归原始发布平台 OpenAI News 所有。